HDR - bizarre sequence of errors

A

alanofcleeves

Member
My HDR has just had a bizarre sequence of problems.
I hope I have remembered them correctly; my efforts may have contributed to the next or obscured the cause - in which case sorry!

The first indication of a problem was when the HDR was brought out of standby to access it via SMB using Win10 / Explorer, something I do almost every day.
After about half a minute the WebIf screen still showed "Initialising ...".
I shut the HDR down, disconnected the power and tried again.
Now the Webif showed a banner "The HDD must be formatted to use recording functions...."

I shut it down again, disconnected the power and tried again.
This time the Humax panel showed that it was in a boot > crash > reboot loop.

I shut it down once more and tried again. This time it booted and Webif warned that bad sectors had been found.
The Disk Diagnostics table seemed to confirm this.
DiskDiag1.jpg

I shut it down once more and now the the WebIf screen just showed that it had crashed.
Crash.log reports that plugins are disabled but gives no clue about what had happened.
Somehow the Pending Sector count has returned to zero and the Reallocated Sector has values unchanged from years ago.
DiskDiag2.jpg

Everything now seemed to work normally, including making a couple of scheduled recordings.
That night I put it into Maintenance Mode and ran fix-disk - no error found or fixed.
[Disk is a Seagate ST2000VM003 with just under 41k power-on hours.]

Currently I am copying saved recordings in case I don't get another chance.
Copying will take a long time but it isn't convenient to take the disk out and copy it offline.

Assuming I can complete this are there any checks I should do?
In particular are there clues in any of the above that point to either software or hardware as the main cause?
Thanks in advance for any insights.
 
Last edited:
At a guess:
  • there is some marginal area on the disk, from which 8 bad sectors have already been patched out;
  • another 8 sectors, probably in the same or a connected (in the direction of rotation, or radially) area, were detected as potentially bad;
  • on the next attempt the potentially bad sectors behaved correctly and were restored to goodness without being patched out.
 
That all sounds plausible, thanks.
Assuming you are correct it might have been safer for ongoing reliability if the additional potentially bad sectors had been patched out rather than restore them but I guess there is no way to force that.

In your opinion: would / could that account for the events leading up to the disk error warnings,
i.e. "The HDD must be formatted to use recording functions...." and the subsequent boot > crash > reboot loop ?
 
You will need to run the fix-flash-packages diagnostic, unless you've previously disabled the auto plugin disabling feature (using the plugin_autodisable/off diagnostic or equivalent).
 
Last edited:
prpr said:
You will need to run the fix-flash-packages diagnostic, unless you've previously disabled the auto plugin disabling feature (using the plugin_autodisable/off diagnostic or equivalent).
Click to expand...
Thanks, I forgot to say that I have done that.

Update:
The box has just crashed again c.8 hours into copying a large folder.
I noted the file involved and have managed to copy it after a restart (much more slowly than the usual 10MB/sec).

I'll persist with the copying and see it I can complete it.
Afterwards would it be worth running a more detailed fix-disk?
 
Agree with ML.

In your opinion: would / could that account for the events leading up to the disk error warnings, i.e. "The HDD must be formatted to use recording functions...." and the subsequent boot > crash > reboot loop ?
Click to expand...

Good point. If the disk errors occurred when the Humax software is trying to validate or mount the disk, it might then report the quoted error. Or if the disk was mounted read-only because of the errors, the same again.

Try fixdisk with the -l (long SMART check) option.
 
I'm impressed that fixdisk knows exactly how long the long test will take (264 minutes).

The mysteries of the message "HDD must be formatted to use recording functions...." followed by a boot > crash > reboot loop remain.

Long and short fixdisk sessions returned zero errors.

DiskDiag1218a.jpg

DiskDiag1218b.jpg

As far as I can tell the HDR is working normally now so I may never know the cause.
Any thoughts, anyone?

I not then many thanks for help and advice with this!
 
alanofcleeves said:
I'm impressed that fixdisk knows exactly how long the long test will take (264 minutes).
Click to expand...
That's something retrieved from the HDD itself, as the test is delegated to the built-in hardware tests.

alanofcleeves said:
The mysteries of the message "HDD must be formatted to use recording functions...." followed by a boot > crash > reboot loop remain.
Click to expand...
Nothing showed up in the file system scans?
 
I've returned the HDR to 'normal' duties - scheduled recordings with a nightly wake-up reminder window for auto processing.
Touch wood, all seems fine, so the mysteries remain.
It will be well-used over the holiday so fingers crossed.

Thanks again for helpful pointers. This forum is amazing.
 
You must log in or register to reply here.
Back
Top